At present, on a WDM (wavelength division multiplexing) network, the spectrum width of a signal is enlarged as a line rate increases. For example, the spectrum width of a 400 Gbit/s signal is generally predicted to exceed 50 GHz. On existing WDM networks, most OADMs (optical add/drop multiplexer) or ROADMs (reconfigurable optical add/drop multiplexer) are set with a 50 GHz spectrum spacing. Therefore, it is generally considered that a 400 Gbit/s or higher-rate signal cannot pass an OADM or ROADM on an existing network.
To solve the foregoing problem, an OADM or ROADM can use the Flex Grid technology. The Flex Grid technology (also referred to as the SLICE or flexible-bandwidth network technology) has become a research hotspot in the industry. The core of this technology is to change a currently fixed spectrum grid (or wavelength spacing, with reference made to ITU-T G.694) into a flexible spectrum grid; that is, a channel of signal may occupy multiple consecutive spectrum grids.
After adopting the Flex Grid technology, the OADM or ROADM uses a flexible-rate/flexible-spectrum-width optical module (transponder) to further improve the fiber spectrum utilization. Using support for OFDM (orthogonal frequency division multiplexing) as an example, the optical module is capable of adjusting the quantity of OFDM subcarriers according to the size of the client-side bandwidth so as to adjust the spectrum width of a line signal and improve the spectrum utilization with help of the control-plane technology.
However, at an OTN (optical transport network, with reference made to the ITU-T G.709 standard) layer, how to encapsulate a client signal, namely, how to implement a flexible-rate OTN signal is a major problem to be solved.
In an existing solution, an OTUk (optical channel transport unit-k, k representing a rate level, and k=1, 2, 3, 4 . . . ) is modulated to a subcarrier. That is to say, a channel of OFDM signal has a data frame format of n×OTUk (n representing the quantity of subcarriers and being an integer greater than or equal to 1).
The existing solution has some problems.
In the existing solution, n OTU/ODU(H) (optical channel data unit (high order)) overheads (with reference made to the ITU-T G.709 standard) are used. Hence, on the management plane, an OTU/ODU(H) overhead needs to be selected as a valid overhead. On an intermediate node, the valid overhead needs to be moved to an OTU/ODU(H) specified by the node. Such processing increases the complexity of management-plane and control-plane design, thereby increasing the difficulty in management and control. In addition, if only one OTU/ODU(H) overhead is selected as the valid overhead, other n−1 OTU/ODU(H) overheads are invalid bytes. In this aspect, the bandwidth is wasted and the bandwidth utilization decreases.